Without using SAN-to-SAN replication, how does data synchronization occur across subnets in a multi-subnet AlwaysOn (AVG) configuration? I want to setup a separate VLAN (replica lan) to synchronize data between three nodes (2 sync in DC1, and 1 async in DC2), but concerned there will need to be additional routing resources to make that possible.
There is no easy answer for this, you will have to go through the entire process from scratch and preferably do it several times so you know exactly what to expect, then focus on monitoring it. This is well documented though and you should start with a search engine query to get familiar with it.
Edwin is a resident HA expert in the SQL Server community and he has great articles on AG's. Here is one about your method. It's too long to post here or shrink, but your answer is to google 'multi subnet ag sql server' or something to that effect. That will lead you to one of several posts, starting a multi part series by Edwin: SQL Server 2012 Multi-Subnet Cluster Part 1